Share on FacebookEmail to someoneTweet about this on TwitterGoogle+Share on LinkedInShare on Reddit

When Google rebranded Webmaster Tools as Google Search Console SEOs were going wild over the new Search Analytics report. However, as the resident skeptic I’m not so quick to worship at the feet of Google’s data.

Google Search Console is full of untrustworthy data

 

Rankings Up, Impressions Down

It seems pretty intuitive to say as your rankings increase, so too should your site’s impressions in SERPs. While this may not always be the case, it is generally true. However, often times in GSC I will see something like this:

Rankings Up, Impressions Down

If GSC data is to be believed, there is often a direct negative correlation between rankings and impressions. Huh?

Non-sensical Keyword Level Data

This keyword level report in GSC is a doozy:

Low volume terms have high impressions

Man, this data is all kinds of jacked up. A long  tail phrase like “local residential moving company” that they rank 37.7th for is driving more impressions then “moving company” a head term that they rank 7.9th for? I mean, I guess that could be true right? Wrong:

Head terms have higher search volume then long tail terms

I’m going to just come out and say it; head terms have higher search volume then long tail ones. Here at Local SEO Guide we thrive on controversy.

A quick caveat, it is totally possible that this company is ranking on the 4th page of Google “nationally” for the phrase local residential moving company and that is what is being counted as impressions. While that goes against the Keyword Planner data, I love indulging in thought experiments. If this were the case, then to me this is another reason why this report is unhelpful. For local businesses, national rankings for terms are unlikely to positively impact their revenue. In this instance the business is a single location moving company based in Detroit. If someone searches local residential moving in Los Angeles and wants to move within the city that is not a lead they can convert. So in a significant amount of cases all this does is pad the impressions and create un-actionable data. Also, in a world of increasing localization of search how are we supposed to be able to differentiate “local” & “national” if Google doesn’t do it for us?

Disparities between GA and GSC

Now, it’s not saying anything new that there are disparities between Google Search Console and Google Analytics, they just measure different things. However, it’s possible to isolate something and show just how bad that disparity is. For our clients, we use UTM campaign tracking parameters to track GMB traffic. So lets look at the data for the page with UTM parameters in GSC and compare it against the campaign tracking in GA:

GSC
GSC is wrong

 

According to GSC we are really awful at the local SEO thing.

GA
Oh wait, there is my traffic


Phew. Keep in mind in the GSC screenshot we are viewing the clicks to that page from a SERP, which should correlate pretty closely with sessions in GA. Now, I know there are other ways to get to a website through A GMB page other then clicking on the website button for a business in the local pack but this disparity is massive.

On a semi-related side note, if you use campaign tracking on a GMB URL then please, for the love of god, canonicalize the GMB landing page and set your URL parameters in GSC.

Final Thoughts

Ever since Google took away keyword data from analytics packages people have been clamoring for better keyword and search data. When Google rolled out the Search Analytics report people in the SEO and digital marketing spaces ate it up, almost religiously. We have been so starved for 1st party Google data for so long everyone seemed willing to take whatever bone Google threw them and treat it as truth because we wanted to believe it was. After playing around with this report for several months I am regularly unable to reconcile it with analytics & ranking data and I have to wonder, how many people are making strategic decisions based on this misleading data and what is the benefit to Google of revamping this report?

 


 
 

14 Comments

  • Julie Kosbab  August 26, 2015 at 10:01 am

    I have been heard to tell several of the LSG Enterprise clientele that Search Console data is for entertainment purposes only.

    Reply
  • Liam  August 26, 2015 at 11:09 am

    Ya I can’t really say I use it all that often as I’m with you, the data is borderline useless.

    Reply
  • Dan Leibson
    Dan Leibson  August 26, 2015 at 11:16 am

    The most exciting of all entertainment 😉

    Reply
  • Steve  August 26, 2015 at 11:42 am

    Yeah, in my experience, you really have to be cautious with it. You can see a ton of impressions for words a client ranks 60th on. Makes it so hard to use.

    Plus, you can’t tell whether it came from the local or organic algorithm (and if it is tracking impressions on the local side).

    The UTM tracking in Google profiles is interesting. Does this cause a problem with consistency?

    Reply
  • David Portney  August 26, 2015 at 1:06 pm

    Hey Andrew,

    For quite some time I wrestled with all of this and even thought maybe I just wasn’t smart enough to reconcile data sets and gave up; coincidentally, right after that, my (former employer) Ian Lurie posted on the Portent blog a rant about how useless and irreconcileable the (then GWT) GSC data is. Now we have a revamped report, and the data is still… “highly questionable” to say the least. Really does make one shake their head about why we cannot get reliable data and get smoke-and-mirror answers in the help section etc. Maybe we should ask John Mueller about this in an upcoming hangout.

    Thanks for making time to rant, er, write about this ongoing problem with GSC data.

    Reply
  • Dan Leibson
    Dan Leibson  August 26, 2015 at 1:19 pm

    Hey David,

    Glad you liked the article, btw my name is Dan 🙂

    Steve – No problem with consistency as long as you canonicalize and set URL parameters.

    Reply
  • Dave  August 26, 2015 at 2:53 pm

    Andrew: Glad you wrote this. Nice job. Webmaster tools/Google console–> it really doesn’t matter what it is called. It is dramatically and purposefully terrible data. Lets face it. Always has been.

    Google took away specifics…specifics you could verify against actual log files or compare what was in GA with a different analytics system…just to get a comfortable level of confidence.

    Instead they substituted this bunch of junk, which they control, which is unverifiable.

    Now we have a couple of smb’s for which we have tracked for years…years before google cut out keyword data. We have also run adwords on them…for areas that are somewhat larger than their actual logical regional coverage.

    The Webmaster/Console data is SO SO out of touch with reality.

    WMT/Console data is worse for a local business than for a national effort. As you said above there is NO NO value for smb’s based in Detroit to get impression data from LA, or San Fran, or Denver…or outside areas. Its worse than worthless…its misleading.

    If, as you suggested much of the seo media and different marketers were putting a nice face on it….its simply misleading.

    I suppose there is a single saving grace. If you aren’t running adwords…or bing ads..it does describe relevant keywords.

    But is its data trustworthy as to impressions and clicks??? I doubt it.

    Last time I looked at it for one smb I had twice the activity in organic google traffic than the console showed.

    Worthless. A pity it ever gets any approval.

    Its google manipulating the masses.

    Reply
  • Dave  August 26, 2015 at 3:13 pm

    Hey, Dan. My apologies. I made the same mistake as the person above.

    Darn good job on the article.

    Reply
  • David Portney  August 26, 2015 at 5:38 pm

    Sorry about that Dan!

    Reply
  • Tiggerito  August 26, 2015 at 9:52 pm

    I also use tracking URLs for my G+ Pages. I’m actually playing with an idea that can make them look and work nicer.

    GSC definitely does not show all the clicks that actually happen. For one client GA shows them, the G+ Page insights show them, but zero in GSC.

    But, it’s the best we can get.

    Reply
  • Dave  August 27, 2015 at 1:17 pm

    Dan: One other point relative to the data you showed and that “crazy” reference to “local residential moving companies” as having beaucoup beaucoup impressions.

    HOGWASH!!!! We know that isn’t accurate. We know “head terms” are searched on much more frequently.

    Interestingly, for years and years I get a wee bit of what I believe is pure BUFFALO SHUFFLE BS from adwords…its been that way in each of 3 accts for certain terms.

    Goog Adwords impressions feeds us some complete whacko long tail keyword phrase. With it…it always shows high impressions.

    (now we do “cover those phrases” via broad term.

    We NEVER get clicks on them. They never ever showed up when we were getting organic data.

    So come on…where and why does Google come up with this stuff???

    How frequent is it??? Do they pepper all kinds of sites with this kind of stuff???

    I’d really really like to look at hundreds of sites and see if examples like this show up across WMT data, across adwords impression data. Are these glitches or are they deliberate???

    Reply
  • Dan
    Dan  August 27, 2015 at 3:57 pm

    Stay tuned Dave, I have some stuff coming up about Keyword Planner as well 🙂

    Reply
  • Gail  September 8, 2015 at 1:53 pm

    Nice article, Dan!

    Reply

Leave A Comment

Please enter your name. Please enter an valid email address. Please enter a message.